Morales Feliciano,et v. John A. Rullan

SELYA, Circuit Judge. Like the legendary Phoenix, this class action litigation involving prison conditions in Puerto Rico is seemingly incapable of eternal rest. The iteration now before us focuses on prospective injunctive relief ordered by the district court in 1998 (the centerpiece of which is the proposed privatization of medical and mental health care throughout the correctional system). Despite his predecessor’s enthusiastic assent to that arrangement, the Commonwealth’s Secretary of Health (the Secretary), acting in his official capacity, sought five years later to vacate or terminate the consent decree embodying that relief. 1 Following an evidentiary hearing, the district court rejected that initiative. Morales Feliciano v. Calderón Serra, 300 F.Supp.2d 321 (D.P.R.2004) (Morales TV). The Secretary now attacks this decision on several fronts. He asserts, inter alia, that the court acted ultra vires; that the court’s 1998 order was void ab initio for failure to meet the requirements of section 802 of the Prison Litigation Reform Act (PLRA), Pub.L. No. 104-134, 110 Stat. 1321-66, § 802 (1996), codified at 18 U.S.C. § 3626; and that in all events, the order should be terminated because the district court’s supportive factfinding was clearly erroneous and infected by errors of law. As a fallback, he argues that the district court bungled the evidentiary hearing and that, at a bare minimum, the matter should be remanded for reconsideration. Finding these importunings unpersuasive, we affirm the decision below. I. BACKGROUND Given the long and tortuous history of this litigation — two years ago, we acknowledged that “the lore of this case is Byzantine,” Morales Feliciano v. Rullán, 303 F.3d 1, 3 (1st Cir.2002) (Morales III) —we confine our introductory comments to those events that are most directly germane to this appeal. The shelves of any well-stocked law library will satisfy the appetites of readers who hunger for more exegetie detail. See, e.g., Morales III, 303 F.3d 1; Morales-Feliciano v. Parole Bd., 887 F.2d 1 (1st Cir.1989); Morales IV, 300 F.Supp.2d 321; Morales Feliciano v. Rosselló Gonzalez, 124 F.Supp.2d 774 (D.P.R.2000); Morales Feliciano v. Rosselló Gonzalez, 13 F.Supp.2d 151 (D.P.R.1998) (Morales II); Morales Feliciano v. Hernandez Colon, 775 F.Supp. 487 (D.P.R.1991); Morales Feliciano v. Hernandez Colon, 775 F.Supp. 477 (D.P.R.1991); Morales Feliciano v. Hernandez Colon, 771 F.Supp. 11 (D.P.R.1991); Morales Feliciano v. Hernandez Colon, 754 F.Supp. 942 (D.P.R.1991); Morales Feliciano v. Hernandez Colon, 704 F.Supp. 16 (D.P.R.1988); Morales Feliciano v. Hernandez *46 Colon, 697 F.Supp. 51 (D.P.R.1988); Morales Feliciano v. Hernandez Colon, 672 F.Supp. 627 (D.P.R.1987); Morales Feliciano v. Hernandez Colon, 697 F.Supp. 26 (D.P.R.1986); Morales Feliciano v. Romero Barcelo, 672 F.Supp. 591 (D.P.R.1986); Morales Feliciano v. Romero Barcelo, 605 F.Supp. 967 (D.P.R.1985); Morales Feliciano v. Romero Barcelo, 497 F.Supp. 14 (D.P.R.1979) {Morales I). In the 1970s, a group of prison inmates initiated a class action alleging “dire shortcomings in virtually every aspect of prisoner confinement” in the Puerto Rican correctional system. Morales III, 303 F.3d at 3. The district court found the plaintiffs likely to succeed on the merits and issued preliminary injunctive relief. Morales I, 497 F.Supp. at 39-41. The defendants made little progress and the district court, in March of 1986, appointed a monitor. See Morales III, 303 F.3d at 3; Morales Feliciano v. Romero Barcelo, 672 F.Supp. at 621. The court charged the monitor with studying various elements of the correctional system and recommending remedial action. With the monitor’s assistance, the court became more proactive and issued a torrent of orders, including temporary restrainers and contempt citations. The court thereafter imposed multimillion dollar fines for the most egregious failures to comply with its directives. In 1990, the court ordered the implementation of medical and mental health care plans recommended by the monitor. These plans transferred the primary responsibility for medical and mental health care in the correctional system from the Administrator of Corrections (the AOC) to the Secretary. The plans also required that the Secretary employ for a minimum of three years a chief health care coordinator (the CHCC) who would bear responsibility for easing the transition and coordinating compliance. Over time, prison conditions improved. In 1996, the district court entered a partial final judgment that settled several disputed issues and urged attempts at consensus-based compliance efforts as to other issues. Despite improvements in many areas, the delivery of medical and mental health care lagged behind. In April of 1997, a court-appointed expert found the existing health care programs incapable of meeting constitutional standards and suggested, as an alternative, the appointment of a receiver. The parties (who, over a quarter of a century, have agreed on little else) unanimously opposed this recommendation. They suggested instead the formation of a private non-profit corporation to provide medical and mental health services to the inmate population. On September 26, 1997, the parties drafted, executed, and filed a stipulation embodying this consensus. Under the terms of the stipulation, a non-profit corporation was to be formed. The corporation would be structured to provide health care services, consistent with the 1990 plan, to all individuals held in institutions operated by the AOC. The stipulation pledged the parties’ full cooperation, required the Secretary and the AOC to lay the groundwork for a transition, and offered to subject the process of privatization to the supervision of the district court. Finally, the stipulation memorialized the parties’ agreement to engage in further discussions concerning the role and authority of the CHCC. The district court endorsed the stipulation and, thus, assumed ownership of the privatization plan that had been conceived by the parties. See Morales III, 303 F.3d at 3-4; see also Morales II, 13 F.Supp.2d at 212-14 (justifying the adoption of the privatization remedy by elaborating on the grave constitutional deficiencies that continued to haunt the existing programs). The order, as it pertains to this appeal, has *47 retained its original substance. The nonprofit entity, known as the Correctional Health Services Corporation (the CHSC), has been formed. The parties and the district court envisioned the CHSC as a key piece of the privatization machinery. Withal, the CHSC was intended as a transitional device rather than as the exclusive provider of inmate health care services in perpetuity. The parties expected that within a few years of its full implementation, the CHSC would compete with private health care providers for contracts to furnish health care services to those persons held in the AOC’s custody. This expectation remains in force. The CHSC has a checkered history. Originally, the parties hoped that it would begin to function as early as July 1, 1998, and become fully operational as a provider of health care to the inmate population by the end of that year. That prediction proved wildly optimistic. Although the district court, using accrued fines, has made substantial funding available to the CHSC — as of the date of the district court’s decision, roughly $55,000,000 had been spent on the privatization solution— this monetary infusion has not yet brought the project to fruition. While the CHSC has made substantial progress toward reaching its stated goals, the pace has been much slower than anticipated. To this point, the CHSC has developed an administrative infrastructure, fashioned an impressive set of substantive programs, and constructed needed facilities. As a result, it has begun contributing to the management of payroll, staff assignments, inventory, purchasing, billing, and financing. The CHSC’s achievements in formulating substantive programs include the creation of an electronic database for health records, patient tracking, and the keeping of appointments; the design of a telemedicine program to enable remote diagnosis and treatment of psychiatric illnesses; and the training of personnel to maximize the efficacy of these (and other) programs. The CHSC also has improved the quality of the facilities and equipment dedicated to the provision of inmate health care. In the same vein, it has purchased and installed computers at most of the institutions operated by the AOC. Last — • but far from least — it is in the process of developing a new acute care hospital. Without in any way minimizing either the value or extent of these accomplishments, it is evident that the CHSC has lagged far behind the timetable that its proponents originally envisioned. Part of the reason is that court supervision has made the process cumbersome (for example, lengthy periods of time have been consumed in the submission and approval of budget proposals). Other factors, such as snail’s-pace negotiations with outside contractors, false starts, and resistance to change, have contributed to the delay. Though this lag time may have been unavoidable, the fact remains that, to this date, not a single patient has been treated by the CHSC. On October 1, 2008, the Secretary, emphasizing this protracted period of delay and the mounting costs of completing the necessary infrastructure, filed a motion under the PLRA to vacate or terminate the privatization component of the extant consent decree. The plaintiffs cross-moved to modify the medical and mental health plans by terminating sections of the decree that assigned the primary responsibility for inmate health care to the Secretary. See Morales IV, 300 F.Supp.2d at 342-43. The cross-motion went unopposed and the district court granted it. Id. at 343. Following an evidentiary hearing on the main motion, the court found that pervasive and persistent constitutional violations *48 continued to shackle the delivery of health care in Puerto Rico’s correctional system. Id. at 323-31. The court also found that the contemplated relief — privatization— satisfied the requirements of the PLRA. Id. at 331-38. Accordingly, the court upheld the challenged portion of the consent decree and directed the AOC to meet with the plaintiff class and submit a further plan for achieving constitutional compliance. Id. at 344-45. The Secretary now appeals. 2 II. APPELLATE JURISDICTION We start by clarifying the extent of our jurisdiction. It is beyond serious question that we have jurisdiction over the main elements of this appeal. See 28 U.S.C. § 1292(a)(1) (conferring jurisdiction over interlocutory orders refusing to terminate injunctions). The Secretary, however, has also attempted to raise a tangential issue. With regard to that issue, appellate jurisdiction is more problematic. The circumstances are as follows. As part of its ruling below, the district court ordered the AOC to “meet [with the plaintiff class] and file with the Court within forty-five (45) days ... a plan on how the [AOC] will assume his legal responsibility to provide health care to inmates.” Morals IV, 300 F.Supp.2d at 345. The Secretary asserts that this order, ex proprio vigore, violates the PLRA and contravenes Supreme Court precedent regarding limitations on the power of federal courts to enforce state law against state officers. Pennhurst State Sch. & Hosp. v. Halderman, 465 U.S. 89, 117, 104 S.Ct. 900, 79 L.Ed.2d 67 (1984) (“[A] federal suit against state officials on the basis of state law contravenes the Eleventh Amendment when ... the relief sought and ordered has an impact directly on the State itself.”). We cannot reach the merits of these claims for we have no jurisdiction over them. This result is largely dictated by the rationale of our decision in a prior phase of this seemingly endless litigation. In Morales III, we ruled that an order that “was within the lower court’s reserved power to establish procedures for compliance with the court’s earlier decrees” was not immediately appealable. 303 F.3d at 10. That reasoning is apposite here. A necessary correlate of the lower court’s duty to supervise the ongoing process of privatization is the power to issue subsidiary orders in aid of that objective. We thus have no jurisdiction to entertain an interlocutory appeal from this portion of the court’s decision (or even to consider the plaintiffs’ contention that the Secretary lacks standing to register such a protest). 3 III. THE MERITS We proceed to examine the remainder of the Secretary’s asseverational array. He has loosed a scattershot attack. Only four of his claims are worthy of extended discussion, namely, (i) that the lower court lacked general equitable power to grant a privatization remedy; (ii) that the PLRA mandates vacation of the 1998 order as void ab initio; (iii) that, after almost five years of wheel-spinning, the PLRA now *49 mandates termination of the prospective relief; and (iv) that the court repeatedly erred in the course of the evidentiary hearing. 4 We address these issues (some of which have subparts) sequentially. A. General Equitable Powers. The Secretary’s first broadside need not detain us. He posits that the relief ordered here is an invalid exercise of federal equitable power. Because this argument presents a purely legal issue, it engenders de novo review. See Sunshine Dev., Inc. v. FDIC, 33 F.3d 106, 111 (1st Cir.1994). At the outset, we note that this argument quite likely is by the boards. The Secretary’s predecessor in office stipulated to the privatization remedy, and there is no claim that he lacked authority to do so. Normally, that would constitute a waiver, which would prevent the Secretary from making this argument altogether. Even if the Secretary is not bound by his predecessor’s acquiescence — a matter on which we take no view — the Secretary did not present this argument to the district court and attempts to raise it for the first time on appeal. “It is a bedrock rule that when a party has not presented an argument to the district court, [he] may not unveil it in the court of appeals.” United States v. Slade, 980 F.2d 27, 30 (1st Cir.1992). At the least, then, the argument is forfeited. We nonetheless inquire briefly into the merits, as forfeited arguments (unlike waived arguments) ordinarily warrant review for plain error. Gómez v. Rivera Rodríguez, 344 F.3d 103, 118 (1st Cir.2003). That type of review “entails four showings: (1) that an error occurred (2) which was clear or obvious and which not only (3) affected the [appellant’s] substantial rights, but also (4) seriously impaired the fairness, integrity, or public reputation of judicial proceedings.” United States v. Duarte, 246 F.3d 56, 60 (1st Cir.2001). The Secretary concedes that injunctions — like this one — that have been issued pursuant to 42 U.S.C. § 1983 are limited only by the court’s inherent equitable powers. See, e.g., Giles v. Harris, 189 U.S. 475, 486, 23 S.Ct. 639, 47 L.Ed. 909 (1903) (Holmes, J.). He insists, however, that a federal court’s equitable powers are restricted to the remedies that were available in equity in 1789 (at the time of the passage of the First Judiciary Act). This proposition is a correct statement of the law, see Grupo Mexicano de Desarrollo v. Alliance Bond Fund, Inc., 527 U.S. 308, 318-19, 119 S.Ct. 1961, 144 L.Ed.2d 319 (1999), but the Secretary’s application of it fails plain error review. Specifically, the Secretary characterizes privatization as “a fairly recent en *50 deavor” and claims, therefore, that there is no reasonable basis for holding that court-imposed privatization of duties ordinarily ascribed to the government is within the district court’s armamentarium. Appellant’s Br. at 29. If this crabbed view reflected the state of the law, equitable remedies would be frozen in time and new remedial applications could never be developed to meet contemporary needs. The view, however, is fundamentally at odds with the core principle that equity must evolve over time “in order to meet the requirements of every case, and to satisfy the needs of a progressive social condition in which new primary rights and duties are constantly arising and new kinds of wrongs are constantly committed.” Union Pac. Ry. Co. v. Chicago, Rock Island & Pac. Ry. Co., 163 U.S. 564, 601, 16 S.Ct. 1173, 41 L.Ed. 265 (1896). Consistent with this principle, the Supreme Court has made clear that the inquiry into the availability of equitable relief in a particular case focuses upon whether the general manner of relief sought was available at equity in 1789. See Grupo Mexicano, 527 U.S. at 322, 119 S.Ct. 1961. This inquiry is informed by a recognition that equity is flexible and that the boundaries of permissible relief are broad. Id. Thus, the relevant question is not whether a specific application of relief was available in 1789, but, rather, whether that general type of relief was available. In this case, the answer to that question is affirmative. At bottom, the endorsed anodyne — privatization—calls for relief in the nature of a mandatory injunction, and injunctive relief is a classic equitable remedy. Great-West Life & Annuity Ins. Co. v. Knudson, 534 U.S. 204, 211 n. 1, 122 S.Ct. 708, 151 L.Ed.2d 635 (2002); Mertens v. Hewitt Assocs., 508 U.S. 248, 256, 113 S.Ct. 2063, 124 L.Ed.2d 161 (1993); Griggs v. E.I. Dupont de Nemours & Co., 237 F.3d 371, 384 (4th Cir.2001). A survey of contemporary applications of this rule supports this conclusion. Courts consistently have upheld relief in equity notwithstanding the fact that the particular remedial application was not available in 1789. School busing orders exemplify this point. See, e.g., Swann v. Charlotte-Mecklenburg Bd. of Educ., 402 U.S. 1, 30, 91 S.Ct. 1267, 28 L.Ed.2d 554 (1971); Morgan v. Kerrigan, 530 F.2d 401, 414-15 (1st Cir.1976). If the Secretary’s argument were sound, all such orders would be beyond the equitable power of a federal court. The short of it is that a federal district court, faced with pervasive and persistent constitutional violations, has not only the power but also the duty to render a decree that will, insofar as is possible, return matters to a constitutionally acceptable level. While the privatization remedy constitutes extraordinary relief and can be justified in only the most exceptional circumstances, the nature of the remedy is no different than that of a garden-variety mandatory injunction. Seen in this light, we conclude that the district court did not commit plain error in determining that it had the authority to approve and implement a privatization remedy jointly proposed by the parties. B. Compliance with the PLRA In 1996, Congress enacted the PLRA, partially in an effort to curb the involvement of the federal judiciary in day-to-day prison management. See 141 Cong. Rec. 13,319 (1995) (statement of Sen. Abraham); id. at 14,418 (statement of Sen. Hatch). This ambient intent must guide our interpretation of the statutory text. See Inmates of Suffolk County Jail v. Rouse, 129 F.3d 649, 655 (1st Cir.1997). The PLRA establishes standards for the entry and termination of all prospective relief in civil actions challenging prison *51 conditions. See Miller v. French, 530 U.S. 327, 333, 120 S.Ct. 2246, 147 L.Ed.2d 326 (2000). The term “prospective relief’ encompasses “all relief other than compensatory monetary damages.” 18 U.S.C. § 3626(g)(7). The statute requires that prospective relief “shall extend no further than necessary to correct the violation of the Federal right of a particular plaintiff or plaintiffs.” Id. § 3626(a)(1). It further provides that a federal court “shall not grant or approve any prospective relief unless the court finds that such relief is narrowly drawn, extends no further than necessary to correct the violation of the Federal right, and is the least intrusive means necessary to correct the violation of the Federal right.” Id. A party is entitled to immediate termination of any relief that has been granted in the absence of findings as to these criteria. Id. § 3626(b)(2). Conversely, however, prospective relief “shall not terminate if the court makes written findings based on the record that prospective relief [meets the above requirements].” Id. § 3626(b)(3). This statutory backdrop provides the setting for the Secretary’s next challenge to the propriety of the privatization order. This challenge has three foci, each spawned by the PLRA. We deal with them in order. 1. Validity of the 1998 Order. The Secretary first argues that, when originally entered in 1998, the order approving privatization did not meet the PLRA’s requirements and, thus, “must be set aside without further consideration.” Appellant’s Br. at 31. Because the district court did not make the findings specifically required by the PLRA coincident with the entry of its 1998 order, the Secretary’s thesis runs, the order was void ab initio. 5 The Secretary’s thesis hinges on the premise that when prospective relief does not meet the PLRA’s requirements at the moment of issuance, the decree is void without reference to the statutory termination procedure. On this premise, the termination procedure applies only to (i) prospective relief entered before the effective date of the PLRA and (ii) prospective relief entered after that date which is accompanied by the requisite narrowness-need-intrusiveness findings. Prospective relief entered after the PLRA’s effective date but unaccompanied by narrowness-need-intrusiveness findings is simply void. There is no textual support for the Secretary’s position. He relies on the mandatory language of 18 U.S.C. § 3626(a), which states that “[t]he court shall not grant or approve any prospective relief’ in the absence of specific findings. But this subsection sets forth the standard district courts must follow when determining whether prospective relief not yet ordered is permitted by the PLRA. Gilmore v. California, 220 F.3d 987, 999 (9th Cir.2000) (“If prospective relief has already been granted by a court, § 3626(b) controls.”). Here, however, privatization constitutes the status quo (and the Secretary advanced no PLRA-based argument at the time that privatization was originally approved). *52 The only authority cited by the Secre tary — Rouse—does not support his point of view. He asserts that Rouse drew a distinction between orders issued prior to the enactment of the PLRA and those issued subsequent thereto. That is incorrect. Rouse drew a distinction between “existing federal court orders” and those “not yet obtained.” 129 F.3d at 654. That distinction contradicts the position that the Secretary advocates in this appeal. Accordingly, we hold that the procedure limned in 18 U.S.C. § 3626(b) applies to any existing prospective relief, regardless of when that relief was first ordered. See Miller, 530 U.S. at 333, 120 S.Ct. 2246 (stating that section 3626(b) applies to “existing injunctions”); Harvey v. Schoen, 245 F.3d 718, 720 (8th Cir. 2001) (holding that prospective relief made “in the absence of the required findings” is immediately terminable “regardless of when ordered”); see also Rouse, 129 F.3d at 654 (suggesting that the termination procedures set forth in section 3626(b) apply to all existing federal court orders). Since privatization is required by a preexisting court order, the situation is controlled by subsection (b), not subsection (a). See Rouse, 129 F.3d at 654. That makes a significant difference. Nothing in subsection (b) or in its legislative history speaks of vacating consent decrees but only of terminating them. That linguistic shift has practical consequences. “While terminating a consent decree strips it of future potency, the decree’s past puissance is preserved....” Id. at 662. That gets the grease from the goose. Although 18 U.S.C. § 3626(b) entitles a defendant to termination of existing prospective relief ordered in the absence of the requisite findings, that proviso excepts cases in which the court supportably finds that the particular form of prospective relief remains at the time of the challenge narrowly tailored, necessary, and relatively unintrusive. See id.; see also Guajardo v. Tex. Dep’t of Crim. Justice, 363 F.3d 392, 394 (5th Cir.2004) (per curiam). There is no time limitation present in section 3626(b)(2), nor does its language suggest that it reaches only those orders issued prior to the PLRA’s effective date. To cinch matters, section 3626(b) provides that any order made in the absence of the required findings is subject to “immediate termination.” This word choice is significant. See Rouse, 129 F.3d at 662. The verb “terminate” means “to put an end to” or “to end.” Black’s Law Dictionary 1482 (7th ed., 1999). If Congress had intended that orders issued in the absence of the required findings be void ab initio, it would almost certainly have chosen a phrase such as “void.” To sum up, the PLRA grants the Secretary a right to move for the termination of prospective relief. The PLRA does not, however, confer any right to argue, five years after the fact, that an order should be deemed void ab initio for lack of contemporaneous findings. Therefore, this aspect of the Secretary’s challenge founders. 2. The Constitutional Violations. Upon a motion to terminate prospective injunctive relief under the PLRA, a court may continue the relief only if it supportably finds that there are ongoing constitutional violations. 18 U.S.C. § 3626(b). The district court purposed to make such findings in this instance. See Morales IV, 300 F.Supp.2d at 323-31, 342. The Secretary counters that the record does not show infirmities sufficient to justify ongoing relief and that the district court’s contrary findings are clearly erroneous. An inquiry into whether current prison conditions constitute an ongoing violation *53 of a federal right “comprises a mixed question of fact and law, the answer to which we review along a degree-of-deference continuum, ranging from plenary review for law-dominated questions to dear-error review for fact-dominated questions.” Rouse, 129 F.3d at 661. Here, the question is fact-intensive—the Secretary frontally challenges the lower court’s fact-finding—so our standard of review is deferential. See id; see also Dedham Water Co. v. Cumberland Farms Dairy, Inc., 972 F.2d 453, 457 (1st Cir.1992) (admonishing that appellate courts should respect the trial court’s findings of fact unless convinced, after a careful review of the record, that a mistake has been made). If there are two permissible views of the evidence, the factfinder’s choice between them must be respected. See Fed. Refin. Co. v. Klock, 352 F.3d 16, 29 (1st Cir.2003). The lower court’s factual findings easily pass muster under this deferential standard. We have painstakingly reviewed the record and examined the numerous findings of fact, mindful that a trial court generally is considered the most informed interpreter of its own prior rulings and findings. See Rouse, 129 F.3d at 661. In this instance, the court was very careful to update the record and to consider the ways in which conditions had changed since its original endorsement of the privatization concept. See Morales IV, 300 F.Supp.2d at 323-31. The court painted a compelling, record-rooted picture of how and in what respects health care for inmates in Puerto Rican prisons remains constitutionally deficient. See id. On this grim record, it is simply implausible to suggest that these findings are clearly erroneous. The district court characterized the denial of medical and mental health services as “massive and systematic.” Id. at 324. That characterization seems apt. No useful purpose would be served by rehearsing the court’s findings in minute detail. We offer instead a few of the more glaring examples. As of 2003, one-fourth of all inmates who requested sick call did not get it; only 55% of all ambulatory care appointments actually took place; only 49% of specialist consultations deemed necessary for serious conditions were arranged; as a rule, medically prescribed diets for inmates were habitually ignored; only 31.3% of inmates who had been diagnosed HIV-positive were receiving treatment; and inmate mortality rates were rising. See id. at 323-31. These are but the tip of a particularly unattractive iceberg. Despite the CHSC’s efforts, physical facilities for inmate health care remain inadequate and there is an acute shortage of psychiatric expertise. See id. at 326-27, 331. Basic health care is critical to maintaining a decent quality of life, and Puerto Rico’s correctional system is not delivering a constitutionally acceptable level of health care to its captive population. We acknowledge that in the course of making these findings, the district court largely disregarded the testimony of the Secretary’s expert witness as lacking in credibility. Although the Secretary complains bitterly about this rejection, it is hornbook law that matters of credibility are ordinarily for the court of first instance, not for the appellate court. Laurent v. Ashcroft, 359 F.3d 59, 64 (1st Cir.2004). Unless the record mandates a contrary finding—and this record most assuredly does not—the court of appeals ought not disturb a trial court’s credibility determination. See Anthony v. Sundlun, 952 F.2d 603, 606 (1st Cir.1991); see also 9A Charles Alan Wright & Arthur R. Miller, Federal Practice & Procedure § 2586 (2d ed.1994) (collecting cases). Here, moreover, there is much to commend the *54 district court’s assessment of the witness’s dubious credibility (for example, the witness provided haphazard and inconsistent testimony and, on cross-examination, withdrew or corrected a myriad of statements). The Secretary accurately observes that some noteworthy advances have been made in the delivery of health care to inmates. He overlooks, however, that notwithstanding these improvements, the district court found substantial deficiencies attendant to virtually every aspect of the inmate health care system. Morales IV, 300 F.Supp.2d at 324. However laudable the advances may be, the district court’s supportable finding that constitutional violations persist suffices to satisfy the requirements of the PLRA and to justify a comprehensive injunctive decree. To say more on this subject would be supererogatory. We hold, without serious question, that the district court’s findings and conclusions about the incidence of continuing constitutional violations are adequately anchored in the record. 3. The Narrowness-Need-Infrusiveness Findings. The PLRA mandates “the termination of extant consent decrees altogether unless the district court makes the specific findings that are necessary to keep a particular decree alive.” Rouse, 129 F.3d at 655. In addition to ongoing constitutional violations, this compendium includes a finding that the ordered relief satisfies the statutory narrowness-need-intrusiveness criteria. See 18 U.S.C. § 3626(b). The district court made such findings here. See Morales IV, 300 F.Supp.2d at 342. The Secretary argues that the facts of this case, as found, require termination of the decree because the chosen remedy — privatization—does not meet the narrowness, need, and intrusiveness requirements delineated in the PLRA. Congress plainly intended the PLRA to operate as a mechanism that would decrease federal judicial involvement in prison administration. See Rouse, 129 F.3d at 655. Congress left room, however, for needed injunctive relief. The Secretary’s insistence that Congress did not want courts involved in prison administration begs the real question- — which is whether a particular set of facts, measured under the statutorily specified criteria, warrants continued judicial involvement. We think that this is a close call in the case at hand, but for the time being, we find supportable the district court’s conclusions regarding narrowness, need, and lack of intrusiveness. The narrowness-need-intrusiveness criteria are, to some extent, self-explicating. The devil is in the details attendant to their application. The application of those criteria is case-specific and must be undertaken in light of both the magnitude of existing constitutional violations and the available remedial alternatives. 6 See, e.g., Clement v. Cal. Dep’t of Corrections, 364 F.3d 1148, 1153 (9th Cir.2004). The constitutional violations here are substantial in both scope and degree. They have defied correction for more than two decades. The district court has tried more conventional measures, but found them wanting. It has afforded the Commonwealth ample opportunity to bring preexisting mechanisms up to speed or otherwise to correct *55 the phalanx of problems. It has witnessed the Commonwealth’s continued inability to cure the constitutional infirmities plaguing the delivery of health care in the correctional system. This record of abject failure matters in the narrowness-need-intrusiveness inquiry. See Benjamin v. Fraser, 343 F.3d 35, 49 (2d Cir.2003) (upholding narrowness-need-intrusiveness findings “in light of the district court’s finding that the [defendant’s] compliance with its remedial responsibilities has been consistently incomplete and inadequate”). The need for continued prospective in-junctive relief is patent. The lower court has been dealing with the deficiencies of Puerto Rico’s correctional system since 1979, and we acknowledge that the sheer passage of so much time counsels restraint in ordering further extensions of prospective injunctive relief. See, e.g., Spangler v. Pasadena City Bd. of Educ., 611 F.2d 1239, 1241 (9th Cir.1979). However, the correctional system is sprawling, the problems with which the court has had to grapple are massive, the Commonwealth’s acceptance of the need for reforms has ranged from inconsistent to. grudging (witness the current state of affairs in which the AOC embraces privatization as the Secretary seeks to dismantle it), and progress has been correspondingly slow. Insofar as health care is concerned, the level of improvement still falls well short of bringing serious violations into constitutional compliance. Morales IV, 300 F.Supp.2d at 323-31. In some respects, backsliding has occurred. See id. at 324. In light of these facts, we have no difficulty affirming the district court’s finding that a need for ongoing injunctive relief exists. We next consider the district court’s findings regarding narrow tailoring and lack of intrusiveness. In 2003, the district court was keenly aware of its duty to ensure that prospective remediation complies with the PLRA’s requirements. See Morales IV, 300 F.Supp.2d at 331. It was, however, faced with a constitutional crisis and only three proposed solutions: turning back the health care clock to reinstate the system that had failed miserably for two decades, placing the system into receivership, or persisting with the plan for privatization (which the parties had conceived and to which they originally had agreed). The status quo ante alternative had proven to be ineffective, and, thus, was undesirable. No one favored the receivership alternative. That left privatization. Against this mise-én-scéne, we cannot say that the lower court’s choice of that alternative was unreasonable. The Secretary nonetheless argues that privatization is much broader than necessary to forfend against cruel and unusual treatment of inmates. If this were the district court’s first attempt at remediation, we quite likely would agree. But the district court has attempted narrower, less intrusive alternatives — and those alternatives have failed to restore the constitutional balance. For this reason, a more innovative remedy is justifiable. The increased intrusiveness and broader scope of the privatization remedy is a direct response to the unique need created by the Commonwealth’s own failure — for more than twenty years — to correct serious constitutional inadequacies. Drastic times call for drastic measures. The Secretary has one final shot in his sling. He laments that privatization is far behind schedule and has accomplished very little to date. Nevertheless, a great deal of time, effort, and money has been spent in constructing facilities, procuring equipment, creating programs, and building a more responsive infrastructure. Perhaps more important, the CHSC has developed a functional information technology system. Knowledge is power, and *56 that platform seems reasonably likely to contribute measurably to the long-term success of the prison health care system. It provides a means of maintaining accurate records and thus creating accountability. See, e.g., Carol Gentry, Health-Care Firms Face Costly Change, Wall St. J., Jan. 3, 2000, at A3 (“By speeding up the transmission and improving the accuracy of information, health plans should be able to do a better job of monitoring the quality of care.”); Laura Landro, Health Care Goes Digital: Doctors and Hospitals Find They Can’t Stay Offline Any Longer, Wall St. J., June 10, 2002, at R6 (“There is a clear linkage now between technology and better patient outcomes.”). The district court has cultivated this tree patiently and at great expense, and it would be rash for us to insist that it be uprooted just when it shows promise of bearing fruit. The Secretary’s related complaint — that the ineffectiveness of privatization to date defeats the finding of narrowness — is unpersuasive for two reasons. First, a determination of narrowness requires a court to decide whether ordered relief is tailored to rectify existing constitutional violations. See Cason v. Seckinger, 231 F.3d 777, 784 (11th Cir.2000) (“The court must make new findings about whether the relief currently complies with the need-narrowness-intrusiveness requirements, given the nature of the current violations.”); see also Castillo v. Cameron County, 238 F.3d 339, 354 (5th Cir.2001). Whether a remedy is capable of successfully ameliorating constitutional violations has no necessary correlation with whether it is narrowly drawn. Narrow relief can be completely ineffectual. See Morales II, 13 F.Supp.2d at 157-58 (noting the failure of limited injunctive orders to bring the Puerto Rican correctional system into constitutional compli-anee). That the privatization concept has been slow to come to fruition does nothing to prove that the order is overly broad (or overly intrusive, for that matter). Second, we think that the PLRA is too blunt an instrument for addressing this concern. As said, the PLRA limits courts to terminating prospective relief. Not every unforeseen difficulty in implementing injunctive decrees necessitates the total abandonment of a remedy. It is true that the CHSC has lagged far behind the original schedule and that the corporation has, to date, not accomplished its ultimate goals. It is also true that an ineffective remedy is a cause for concern. In light of the substantial progress that has been made, however, the mere fact that privatization is behind schedule does not convince us that it fails the narrowness-need-intrusiveness test. 7 We are well aware that federal oversight of Puerto Rico’s correctional system cannot — and should not — last forever. We strongly urge both the district court and the litigants to move the privatization process forward with all practicable speed. For now, however, we conclude, based on the record evidence and the factual findings of the district court, that the privatization order meets the PLRA’s requirements. The privatization solution has not yet been put into practice and common sense dictates that it be given a fair chance to work. The order appealed from provides that chance. C. The Evidentiary Hearing. The Secretary’s fallback position is that the district court committed reversible error in its manner of conducting the eviden-tiary hearing on the motion to terminate *57 prospective relief. This contention has four branches. All of them are acarpous. 1. Burden of Proof. The Secretary first suggests that the district court imper-missibly shifted the burden of proof. This suggestion derives primarily from the court’s decision to order the proof by directing the Secretary (rather than the inmate class) to come forward, at the start of the hearing, with his evidence. It is axiomatic that district courts enjoy wide latitude in matters concerning the ordering of proof and the presentation of evidence. See, e.g., United States v. Holmquist, 36 F.3d 154, 163 (1st Cir.1994); Elgabri v. Lekas, 964 F.2d 1255, 1260 (1st Cir.1992); see also Fed.R.Evid. 611(a). We review a trial court’s determinations concerning the mode and order of proof for abuse of discretion. Nat’l R.R. Pass. Corp. v. Certain Temporary Easements, 357 F.3d 36, 42 (1st Cir.2004). We will reverse only if a determination has unfairly prejudiced the complaining party. Id. The decision to require the Secretary to present his proof first was not an abuse of discretion. In arriving at its decision on the motion to terminate, the court was free to rely on any aspect of the record that tended to shed light on whether constitutional violations persisted. See Laaman v. Warden, N.H. State Prison, 238 F.3d 14, 17-18 (1st Cir.2001) (noting that the court’s familiarity with the record and its receipt of periodic reports may afford it “comprehensive knowledge of whether a prison is, or is not, continuing to violate the consent decree”). At the beginning of the hearing, the court had before it a full evidentiary record, developed over many years, together with a cache of periodic reports as to progress under the consent decree. This plethoric evidence was sufficient, if unrebutted, to support a finding that constitutional violations endured. With this in mind, it was sensible for the court to invite the Secretary to offer evidence to the contrary. To be sure, a trial court’s authority to regulate the order of proof does not include the power to shift the burden of proof. Brown Daltas & Assocs., Inc. v. Gen. Accid. Ins. Co., 48 F.3d 30, 37 (1st Cir.1995) (reversing a judgment because the trial court erroneously shifted the burden of proof); see also 9 J. Wigmore, Evidence § 2489 (J. Chadbourn ed.1981). But that is not what happened here: the court’s election to hear the Secretary’s evidence first in no way betokened reallocation of the burden of proof. In its subsequent review of the evidence, the court said nothing that indicated any misunderstanding as to where the burden of proof rested. Rather, it supportably determined that the facts presented by the Secretary lacked sufficient probative force to blunt the thrust of the accumulated evidence of record. Morales IV, 300 F.Supp.2d at 339-40 (relying on the Secretary’s own evidence to find continuing constitutional violations). In short, the contention that the lower court impermissibly shifted the burden of proof has no footing in the record. 2. The Agreement Between the Parties. The Secretary next maintains that the district court erred by relying on the existence of a “private agreement” (by private agreement, he means the stipulation that was signed in 1997 by, inter alios, his predecessor in office). Specifically, the Secretary asserts that the fact that there was an agreement between the parties should have had no bearing on the continuing viability of the subsequent order approving privatization (entered in 1998). This assertion overlooks that the so-called private agreement was embodied in a stipulation presented to and adopted by the district court. Upon its adoption, the pri *58 vate agreement became a judicially enforceable consent decree. Frew ex rel. Frew v. Hawkins, 540 U.S. 431, 124 S.Ct. 899, 903, 157 L.Ed.2d 855 (2004) (“A consent decree embodies an agreement of the parties and is also an agreement that the parties desire and expect will be reflected in, and be enforceable as, a judicial decree that is subject to-the rules generally applicable to other judgments and decrees.” (internal quotation marks omitted)); In re Pearson, 990 F.2d 653, 658 (1st Cir.1993) (similar). This is significant because the Secretary’s argument blurs the crucial distinction, for PLRA purposes, between private side agreements and consent decrees. The latter may serve as a vehicle for prospective relief under the PLRA; the former may not. See 18 U.S.C. § 3626(a); see also id. § 3626(c)(1). In all events, the lower court did not rely impermissibly on the parties’ original agreement. Although the court noted the existence of that agreement and scolded the Secretary for trying to renege on it, the holding in Morales IV was in no way premised either on an estoppel or on the fact that the parties previously had agreed to explore a privatization alternative. For these reasons, the Secretary’s argument fails. 3. The Exclusion of Evidence. The Secretary also contends that the district court erred in refusing to allow him to introduce evidence as to the narrowness-need-intrusiveness criteria. On this point, as is so often true in litigation, context is decisive. After the Secretary had presented his evidence as to the existence vel non of continuing constitutional violations, the district court asked him to summarize the evidence that he planned to adduce on the narrowness-need-intrusiveness criteria. The Secretary proceeded to make an offer of proof. Having heard the proffer, the court determined that nothing in it bore directly upon the narrowness of, need for, or potential intrusiveness of the privatization remedy. Consequently, the court rejected the Secretary’s offer of proof on relevancy grounds. We descry no error. In many instances, relevancy boils down to a judgment call. See generally Fed.R.Evid. 401 (deeming evidence relevant if it has “any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence”). Accordingly, district courts enjoy wide latitude in determining the relevancy vel non of evidence. Laaman, 238 F.3d at 18-19; United States v. Tierney, 760 F.2d 382, 387 (1st Cir.1985). In this instance, the district court concluded that the offer of proof had no bearing on the appropriateness of the remedy under the PLRA’s narrowness-need-intrusiveness criteria. Morales IV, 300 F.Supp.2d at 337-38. Having reviewed the offer of proof, we find that this ruling was within the realm of the district court’s discretion. We explain briefly. By way of his offer of proof, the Secretary proposed to introduce evidence designed to show the ineffectiveness of the remedy. He began the proffer by alleging generally that the non-profit corporation was not a viable remedy because it had not yet treated a single patient. The district court asked for elaboration, and the Secretary expressed a desire to take discovery on the matter. Morales IV, 300 F.Supp.2d at 337. When the court refused that untimely request, the only details proffered by the Secretary tended to show that the CHSC was not complying with either the district court orders or Puerto Rico laws, and that it was not ready to begin treating patients. Id. at 338. This is a distinct line *59 of inquiry. See Hadix v. Johnson, 228 F.3d 662, 673 (6th Cir.2000) (explaining that inquiries into compliance and inquiries into necessity under the PLRA are separate and distinct). As the district court sagely noted, compliance was not at issue in the evidentiary hearing. Morales IV, 300 F.Supp.2d at 338. We — like the district court — do not see how the Secretary’s offer of proof informs a determination of whether the privatization remedy is a narrowly tailored, relatively unintrusive response to the serious constitutional violations that the district court found still existed. We note, moreover, that the Secretary has not explained the connection. We thus cannot fault the district court either for the rejection of the offer of proof or for its consequent exclusion of the proffered evidence. 4. Rule 52(c). Finally, the Secretary assigns error to the district court’s use of Fed.R.Civ.P. 52(c) as an adjudicative tool. See Morales IV, 300 F.Supp.2d at 344. Rule 52(c) provides a court conducting a bench trial with a means for issuing a judgment on partial findings. The rule is designed for use when “a party has been fully heard on an issue and the court finds against the party on that issue.” Fed.R.Civ.P. 52(c). The district court invoked that procedure here. The Secretary asseverates that this was not permissible on the facts of this case. We reject this asseveration. To be frank, we find the Secretary’s argument difficult to fathom. The text of the rule is clear. When a party has finished presenting evidence and that evidence is deemed by the trier insufficient to sustain the party’s position, the court need not waste time, but, rather, may call a halt to the proceedings and enter judgment accordingly. See Fed.R.Civ.P. 52 advisory committee note (“Subdivision (c) ... authorizes the court to enter judgment at any time that it can appropriately make a dispositive finding of fact on the evidence.”); see also Atl. Track & Turnout Co. v. Perini Corp., 989 F.2d 541 (1st Cir.1993). The fact that the inmate class had the burden of proof did not place Rule 52(c) off limits. A judgment under Rule 52(c) can be entered for or against either a plaintiff or a defendant, regardless of the allocation of the burden of proof. See Bursztajn v. United States, 367 F.3d 485, 488 (5th Cir.2004); Fed. Refin. Co., 352 F.3d at 26; see also 9A Wright & Miller, supra § 2573.1. Here, the Secretary had advanced his proof and the district court supportably concluded that the Secretary’s evidence could not withstand the probative force of the historical record. In other words, even after taking the Secretary’s case into account, the court determined that the plaintiffs had sustained their burden of showing pervasive and persistent constitutional violations. Morales IV, 300 F.Supp.2d at 340-42. The court also sup-portably determined that the privatization remedy satisfied the narrowness-need-intrusiveness criteria, and that the Secretary’s offer of proof did not include any relevant evidence to the contrary. Based on these findings, the district court appropriately entered judgment pursuant to Rule 52(c). That the court made a series of credibility calls in reaching these conclusions in no way detracts from the legitimacy of its approach. See Wright & Miller, supra § 2573.1, at 497-99 (noting that the court’s task under Rule 52(c) is to weigh the evidence, without drawing any special inferences in the nonmovant’s favor, resolve any conflicts in the evidence, and “decide for itself where the preponderance lies”). IV. CONCLUSION We are sensitive both to Congress’s manifest desire to limit federal judicial oversight of state correctional systems and to the length of time that Puerto *60 Rico’s prisons have been under federal hegemony. These concerns impel us to counsel the parties and the district court to move forward as swiftly as possible to cross the threshold of constitutional adequacy so that federal judicial involvement may cease. For now, however, we accept the well-documented judgment of the district court — a court intimately acquainted with the details of this litigation and the intricacies of the Puerto Rican correctional system — that inmate health care remains constitutionally unacceptable and that the privatization remedy holds promise for bringing inmate health care into compliance with constitutionally mandated standards. 8 Consequently, we reject the Secretary’s current effort to vacate or terminate the existing consent decree. We add a few words of caution. There are now significant financial reserves (in the form of accumulated fines) available for improvements in health care delivery, and the AOC seems willing to work for change. We urge the district court to move toward extricating itself from the toils of this litigation as soon as it can do so without defaulting on its responsibilities under the Constitution. We need go no further. For the foregoing reasons, we affirm the district court’s denial of the Secretary’s motion to vacate or terminate the pertinent portions of the existing consent decree. We take no view of the district court’s procedural order, see supra Part II, as we lack jurisdiction to review that order on this interlocutory appeal. Affirmed. Costs shall be taxed in favor of the plaintiffs. Footnotes [1] We note that a new administration, representing a different political party, took office following the general election held in November of 2000. [2] The AOC has not appealed, and the district court's ukase has not been stayed. We are advised that meetings between the AOC and the plaintiffs are proceeding apace. [3] With only isthmian exceptions, the courts of appeals are prohibited from exercising pendent appellate jurisdiction. See Swint v. Chambers County Comm’n, 514 U.S. 35, 49-51, 115 S.Ct. 1203, 131 L.Ed.2d 60 (1995); Limone v. Condon, 372 F.3d 39, 51-52 (1st Cir.2004). The Secretary offers no developed argumentation on this point and, as matters stand, we believe that this case is not a credible candidate for the exercise of pendent appellate jurisdiction. [4] The Secretary makes a fifth argument: that the relief ordered by the district court requires Puerto Rican officials to act in contravention of local law. The district court refused to address this argument on the ground that the Secretary did not sufficiently specify the nature of the alleged violations. That ruling is supportable, and it precludes reliance on the argument here. See Teamsters Union v. Superline Transp. Co., 953 F.2d 17, 21 (1st Cir.1991) (holding that arguments not squarely raised in the trial court cannot be advanced on appeal). In all events, the Secretary's appellate presentation suffers from the same lack of specificity; he alleges the violations in the most general terms and does not inform us of where in the record we might find any evidence of them. This lack of developed appellate argumentation constitutes an independently sufficient basis for deeming the argument abandoned. See United States v. Zannino, 895 F.2d 1, 17 (1st Cir.1990) ("It is not enough merely to mention a possible argument in the most skeletal way, leaving the court to do counsel's work, create the ossature for the argument, and put flesh on its bones.”). [5] Although we assume arguendo the correctness of the Secretary's assessment of what transpired in 1998, we note that the district court made a series of contemporaneous written findings, which may well have satisfied the narrowness-need-intrusiveness requirements. See Morales II, 13 F.Supp.2d at 212 (holding that privatization “is necessary to protect the members of the plaintiff class from cruel and unusual punishment through constitutionally unacceptable health care services,” that the remedy "extends no further than is necessary to correct the violation of the Federal rights of members of the plaintiff class,” that "such relief is narrowly drawn,” and that there "is no less intrusive means to correct the violation of plaintiffs' Federal rights”). [6] We reject out of hand the Secretary's argument that an order to privatize is a per se violation of the narrowness requirement of the PLRA. The narrowness of a remedy necessarily involves a close examination of a particular set of facts and a determination concerning what steps are reasonable to cure the ongoing constitutional violations. Because this will necessarily be influenced by a variety of factors, we are wary of adopting ironclad rules. [7] It may well be that modification, or some other more limited adjustment, is appropriate to address such concerns, see Fed.R.Civ.P. 60(b)(5) — but that issue is not before us. [8] Should the circumstances change, or a substantial interval pass without significant progress, the Secretary is, of course, free to ask the lower court to reexamine these findings and reconsider the advisability of the ordered relief.